Ventilation balance with different pressure drop and flow requirements 1

[xj25]

As I don´t see a lot of activity in the ventilation forum, I would ask if someone can
make some comments/suggestions to the following thread,

thanks in anycase

thread1043-416400

 

[Drazen]

i'm curious that cfd can be put to use for the one-branch problem, but in this particular case cfd will not help you, you need to calculate the whole network, not flow in one spot.

what you deem as "having detailed calculation for each branch"? in hvac practice that would mean "having detailed calculation for each flow path", but as it is seemingly not your expertise you may mean something different.

balancing has to be conducted for each and every path on the network, not for one branch only. after all paths are calculated, results need to show which balancing element can be used for which node, based on calculated balancing pressures for paths. you need hvac engineer for all that job.

 

[lilliput1]

See my post in the thread.

 

[xj25]

Drazen (Mechanical) 5 Nov 16 15:49
i'm curious that cfd can be put to use for the one-branch problem, but in this particular case cfd will not help you, you need to calculate the whole network, not flow in one spot.

what you deem as "having detailed calculation for each branch"? in hvac practice that would mean "having detailed calculation for each flow path", but as it is seemingly not your expertise you may mean something different.

balancing has to be conducted for each and every path on the network, not for one branch only. after all paths are calculated, results need to show which balancing element can be used for which node, based on calculated balancing pressures for paths. you need hvac engineer for all that job.

Thanks for the reply.

Our supplier have done CFDs calculation for "branch A" and "branch B", so to get the system curve for each flow path.
Then they have done a calculation of the fan outlet T section, usign as outlet conditions the system curve for each branch.
They have done then the "experiment" of moving the deflector to balance the needed flow sharing, so to get to that weird proposal seen in the image.

I think that a more sensible option is:
1- use a deflector, miter corners or whatever to make the flow turn the T without a big presssure drop
2- use cross section areas based on design flow for branch outlets, to get a reasonable speed/pressure drop/noise.
3- then use a damper restriction in the upper brach (A), to balance the pressure drop to the design flow.

I am not an HVAC expert, but our system supplier is supposed to be, so I would like to have some basic ideas clear before discussing with them.

Thanks.

 

[Drazen]

Again, what does it mean "calculation for each branch"? I really grow tired of repeating myself in these threads, not getting response.

How could your super-professional cfd specialist know what is outlet pressure of the tee without calculating all paths down to the outlets? This tee is not separate from the network, and calculating it without having calculated all interactions with the network does not make any sense.

 

[xj25]

I will try to be more detailed. Let´s see...

We have 3d models for both ducting with all details, outlets etc. for "branch A" and "Branch B".
It has been done a CFD calculation with two defined flow inputs for each branch.
So we have two points of the (Inlet Pressure-Inlet Flow) curve for each branch.
We get the parabola equations correspondig to that.

Now we have also the 3d model for the tee (seen in picture).
With the CFD solver, you can use as boundary condition for the outlets, the Pressure-Flow curves
calculated above. So to simplify the problem and avoid the need to calculate the full model (too big model).
For the "Tee" inlet boundary condition we use a fixed objetive flow.
The solver crunch numbers, and if it converges, it will give you the flow field and the pressure field for all the "Tee" model. It should behave (as an approximation) as if the actual branch A-B are really conected.

That´s all, I hope it is clearer now.

Regards

 

[lilliput1]

You could just use a tee with 45 degree sides on both inner corner (like back to back shoe taps with the inlet opening of each sized for matching intake velocity at eaxh btranch with their corresponding CFM. No need to spend time doing CFD calcs.